Dictionary generation for downhole signal compression

Abstract

An apparatus includes a processor and a machine-readable medium having program code to cause the apparatus to obtain a first dictionary based on a first training set of signals and determine a first subset of the first training set of signals based on a training reconstruction accuracy threshold and the first dictionary, wherein each atom in the first dictionary includes at least one of a signal pattern and a function representing the signal pattern. The program code also includes code to generate a second dictionary based on a second training set of signals, wherein the second training set of signals includes the first subset of the first training set of signals.

What is claimed is: 
     
       1. An apparatus comprising:
 a processor; and 
 a machine-readable medium having program code executable by the processor to cause the apparatus to,
 obtain a first dictionary based on a first training set of signals, wherein each atom in the first dictionary comprises at least one of a signal pattern and a function representing the signal pattern, 
 determine a first subset of the first training set of signals based on a training reconstruction accuracy threshold and the first dictionary, and 
 generate a second dictionary based on a second training set of signals, wherein the second training set of signals comprises the first subset of the first training set of signals. 
 
 
     
     
       2. The apparatus of  claim 1 , further comprising program code to:
 perform a set of reconstructions of a second set of signals using the first dictionary to determine a first set of atoms having a corresponding set of combined atom weights, wherein each atom in the first set of atoms has a corresponding combined atom weight less than a weight threshold; and 
 generate a modified dictionary, wherein generation of the modified dictionary comprises at least one of an operation to generate a third dictionary that does not include the first set of atoms and an operation to update the first dictionary to remove the first set of atoms from the first dictionary. 
 
     
     
       3. The apparatus of  claim 1 , further comprising program code to:
 perform a set of reconstructions of a second set of signals using the second dictionary to determine a first set of atoms having a corresponding set of combined atom weights, 
 wherein each atom in the first set of atoms has a combined atom weight less than a weight threshold; and 
 generate a modified dictionary, wherein generation of the modified dictionary comprises at least one of an operation to generate a third dictionary that does not include the first set of atoms and an operation to update the second dictionary to remove the first set of atoms from the second dictionary. 
 
     
     
       4. The apparatus of  claim 1 , further comprising program code to:
 obtain an independently-determined atom, wherein the independently-determined atom is determined independently of the first training set of signals; and 
 add the independently-determined atom to at least one of the first dictionary and the second dictionary to generate a modified dictionary. 
 
     
     
       5. The apparatus of  claim 4 , wherein first derivative values of the independently-determined atom are continuous. 
     
     
       6. The apparatus of  claim 4 , wherein the program code to add the independently-determined atom further comprises program code to determine whether the independently-determined atom is in at least one of the first dictionary and the second dictionary. 
     
     
       7. The apparatus of  claim 1 , further comprising program code to generate a third dictionary, wherein the third dictionary has a third dictionary signal length that is different from a first dictionary signal length, wherein the first dictionary signal length corresponds with the first dictionary. 
     
     
       8. The apparatus of  claim 1 , further comprising program code to subtract an offset pattern value from at least one of the first training set of signals and the second training set of signals. 
     
     
       9. A method comprising:
 obtaining a first dictionary based on a first training set of signals, wherein each atom in the first dictionary comprises at least one of a signal pattern and a function representing the signal pattern; 
 determining a first subset of the first training set of signals based on a training reconstruction accuracy threshold and the first dictionary; and 
 generating a second dictionary based on a second training set of signals, wherein the second training set of signals comprises the first subset of the first training set of signals. 
 
     
     
       10. The method of  claim 9 , further comprising:
 performing a set of reconstructions of a second set of signals using the first dictionary to determine a first set of atoms having a corresponding set of combined atom weights, wherein each atom in the first set of atoms has a corresponding combined atom weight less than a weight threshold; and 
 generating a modified dictionary, wherein generating the modified dictionary comprises at least one of generating a third dictionary that does not include the first set of atoms and updating the first dictionary to remove the first set of atoms from the first dictionary. 
 
     
     
       11. The method of  claim 9 , wherein the method further comprises:
 performing a set of reconstructions of a second set of signals using the second dictionary to determine a first set of atoms having a corresponding set of combined atom weights, wherein each atom in the first set of atoms has a combined atom weight less than a weight threshold; and 
 generating a modified dictionary, wherein generating the modified dictionary comprises at least one of generating a third dictionary that does not include the first set of atoms and updating the second dictionary to remove the first set of atoms from the second dictionary. 
 
     
     
       12. The method of  claim 9 , wherein the method further comprises:
 obtaining an independently-determined atom, wherein the independently-determined atom is determined independently of the first training set of signals; and 
 adding the independently-determined atom to at least one of the first dictionary and the second dictionary to generate a modified dictionary. 
 
     
     
       13. The method of  claim 9 , further comprising generating a third dictionary, wherein the third dictionary has a third dictionary signal length that is different from a first dictionary signal length, and wherein the first dictionary signal length corresponds with the first dictionary. 
     
     
       14. The method of  claim 9 , further comprising subtracting an offset pattern value from at least one of the first training set of signals and the second training set of signals. 
     
     
       15. One or more non-transitory machine-readable media comprising program code, executable on one or more computing devices, for generating a dictionary, the program code to:
 obtain subsurface measurement series using a subsurface sensor in a borehole; 
 obtain a first dictionary based on a first training set of signals, wherein each atom in the first dictionary comprises at least one of a signal pattern and a function representing the signal pattern; 
 determine a first subset of the first training set of signals based on a training reconstruction accuracy threshold and the first dictionary; and 
 generate a second dictionary based on a second training set of signals, wherein the second training set of signals comprises the first subset of the first training set of signals. 
 
     
     
       16. The one or more non-transitory machine-readable media of  claim 15 , further comprising program code to:
 perform a set of reconstructions of a second set of signals using the first dictionary to determine a first set of atoms having a corresponding set of combined atom weights, wherein each atom in the first set of atoms has a corresponding combined atom weight less than a weight threshold; and 
 generate a modified dictionary, wherein generation of the modified dictionary comprises at least one of an operation to generate a third dictionary that does not include the first set of atoms and an operation to update the first dictionary to remove the first set of atoms from the first dictionary. 
 
     
     
       17. The one or more non-transitory machine-readable media of  claim 15 , further comprising program code to:
 perform a set of reconstructions of a second set of signals using the first dictionary to determine a first set of atoms having a corresponding set of combined atom weights, wherein each atom in the first set of atoms has a corresponding combined atom weight less than a weight threshold; and 
 generate a modified dictionary, wherein generation of the modified dictionary comprises at least one of an operation to generate a third dictionary that does not include the first set of atoms and an operation to update the second dictionary to remove the first set of atoms from the second dictionary. 
 
     
     
       18. The one or more non-transitory machine-readable media of  claim 15 , further comprising program code to:
 obtain an independently-determined atom, wherein the independently-determined atom is determined independently of the first training set of signals; and 
 add the independently-determined atom to at least one of the first dictionary and the second dictionary to generate a modified dictionary. 
 
     
     
       19. The one or more non-transitory machine-readable media of  claim 15 , further comprising program code to generate a third dictionary, wherein the third dictionary has a third dictionary signal length that is different from a first dictionary signal length, wherein the first dictionary signal length corresponds with the first dictionary. 
     
     
       20. The one or more non-transitory machine-readable media of  claim 15 , further comprising program code to subtract an offset pattern value from at least one of the first training set of signals and the second training set of signals.